beit-base-patch16-224-85-fold4

This model is a fine-tuned version of microsoft/beit-base-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.1725
• Accuracy: 0.9545

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	2	0.6874	0.6818
No log	2.0	4	0.7083	0.7045
No log	3.0	6	0.8871	0.7045
No log	4.0	8	0.7305	0.7045
0.6246	5.0	10	0.5791	0.7045
0.6246	6.0	12	0.5888	0.7045
0.6246	7.0	14	0.6050	0.7045
0.6246	8.0	16	0.5468	0.7045
0.6246	9.0	18	0.5351	0.7045
0.4453	10.0	20	0.4155	0.8409
0.4453	11.0	22	0.8266	0.7045
0.4453	12.0	24	0.3905	0.8409
0.4453	13.0	26	0.3942	0.8409
0.4453	14.0	28	0.4015	0.8409
0.3613	15.0	30	0.3474	0.8182
0.3613	16.0	32	0.4763	0.8182
0.3613	17.0	34	0.3894	0.7955
0.3613	18.0	36	0.4290	0.7955
0.3613	19.0	38	0.3525	0.8636
0.2928	20.0	40	0.3426	0.8864
0.2928	21.0	42	0.4060	0.8182
0.2928	22.0	44	0.6962	0.75
0.2928	23.0	46	0.3514	0.8636
0.2928	24.0	48	0.5302	0.8409
0.2256	25.0	50	0.3094	0.8636
0.2256	26.0	52	0.2977	0.8636
0.2256	27.0	54	0.4883	0.8182
0.2256	28.0	56	0.3008	0.8409
0.2256	29.0	58	0.3226	0.8409
0.2231	30.0	60	0.4101	0.8409
0.2231	31.0	62	0.3197	0.8409
0.2231	32.0	64	0.4133	0.7727
0.2231	33.0	66	0.2923	0.8636
0.2231	34.0	68	0.4391	0.8636
0.1756	35.0	70	0.3016	0.8636
0.1756	36.0	72	0.2749	0.9091
0.1756	37.0	74	0.3146	0.8864
0.1756	38.0	76	0.3095	0.8409
0.1756	39.0	78	0.3017	0.8864
0.1592	40.0	80	0.2762	0.8864
0.1592	41.0	82	0.4054	0.8409
0.1592	42.0	84	0.2787	0.8864
0.1592	43.0	86	0.3193	0.8636
0.1592	44.0	88	0.2783	0.9091
0.1857	45.0	90	0.2934	0.9091
0.1857	46.0	92	0.3579	0.8636
0.1857	47.0	94	0.3501	0.8864
0.1857	48.0	96	0.3358	0.8864
0.1857	49.0	98	0.2981	0.9091
0.1179	50.0	100	0.3364	0.8636
0.1179	51.0	102	0.3324	0.8636
0.1179	52.0	104	0.1725	0.9545
0.1179	53.0	106	0.1222	0.9545
0.1179	54.0	108	0.1500	0.9091
0.1448	55.0	110	0.2358	0.9091
0.1448	56.0	112	0.2224	0.9091
0.1448	57.0	114	0.1457	0.9318
0.1448	58.0	116	0.1745	0.9318
0.1448	59.0	118	0.1990	0.9091
0.1343	60.0	120	0.2905	0.8864
0.1343	61.0	122	0.3842	0.8864
0.1343	62.0	124	0.3031	0.8864
0.1343	63.0	126	0.2642	0.8864
0.1343	64.0	128	0.2412	0.9091
0.1109	65.0	130	0.3347	0.8864
0.1109	66.0	132	0.4005	0.8864
0.1109	67.0	134	0.2905	0.8864
0.1109	68.0	136	0.3168	0.9318
0.1109	69.0	138	0.3845	0.8864
0.1221	70.0	140	0.3178	0.9318
0.1221	71.0	142	0.2690	0.9318
0.1221	72.0	144	0.2516	0.8864
0.1221	73.0	146	0.2347	0.9091
0.1221	74.0	148	0.2376	0.9318
0.1191	75.0	150	0.2480	0.9318
0.1191	76.0	152	0.2597	0.9091
0.1191	77.0	154	0.3071	0.9091
0.1191	78.0	156	0.3354	0.9091
0.1191	79.0	158	0.2988	0.8864
0.1133	80.0	160	0.2760	0.9091
0.1133	81.0	162	0.2832	0.9318
0.1133	82.0	164	0.2793	0.9318
0.1133	83.0	166	0.2779	0.9091
0.1133	84.0	168	0.3004	0.8864
0.098	85.0	170	0.3275	0.8864
0.098	86.0	172	0.3394	0.8864
0.098	87.0	174	0.3257	0.8864
0.098	88.0	176	0.3172	0.8864
0.098	89.0	178	0.3122	0.9318
0.0917	90.0	180	0.3208	0.9318
0.0917	91.0	182	0.3236	0.9318
0.0917	92.0	184	0.3274	0.9318
0.0917	93.0	186	0.3331	0.8864
0.0917	94.0	188	0.3379	0.8864
0.0989	95.0	190	0.3404	0.8864
0.0989	96.0	192	0.3452	0.8864
0.0989	97.0	194	0.3491	0.8864
0.0989	98.0	196	0.3489	0.8864
0.0989	99.0	198	0.3482	0.8864
0.0916	100.0	200	0.3467	0.8864

Framework versions

• Transformers 4.40.2
• Pytorch 2.2.1+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1